Just How Broken is John Axford, Really?

Monday afternoon, the Brewers were leading the Cubs 7-2 going into the bottom of the ninth. Brandon Kintzler took the mound, but after three straight batters reached base, he was replaced by Jim Henderson. Henderson allowed a little bit of damage, but he successfully slammed the door, picking up a save. Henderson pitched in part because John Axford threw 18 pitches on Sunday. Henderson pitched more because Axford allowed two more runs Sunday, bringing him to a season total of six in 2.2 innings. Fueling those six runs allowed have been four dingers, as Axford’s problems from 2012 appear to have carried over into the new campaign.

The talk now is that Henderson will replace Axford as the Brewers’ closer. Axford has been getting booed at home, on account of the sucking, and if the Brewers want to contend and make the playoffs, they can’t afford to have an unreliable closer who’s demonstrated his unreliability. Many feel that Axford has earned a demotion. Many reached that point ages ago. It’s no longer a question of whether Axford should be demoted. It’s a question of: what’s the matter with John Axford?

A couple relevant tweets, I think:

 


Tipping pitches is a bad idea, since pitching is so much about sequencing and unpredictability. It’s not the first whispers we’ve heard of Axford tipping his pitches, but it’s not an open-and-shut case. What that is is a scout’s assertion, and the scout is biased by the knowledge that Axford has been getting lit up. How much is actually wrong with Axford, and how much are people just trying to look for reasons?

Between 2010-2011, Axford was good. He allowed five home runs in 124 games. Between 2012-2013, Axford has been bad. He’s allowed 14 home runs in 78 games. What I’m going to show you now is a table of stats. Half of these stats were posted by Axford between 2010-2011, and half of them have been posted by Axford between 2012-2013. They are jumbled and unlabeled to make an intentional point.

K% BB% GB/FB FBv (mph) Contact% Zone%
30% 12% 1.4 96.1 75% 47%
30% 10% 1.4 95.3 76% 48%

Something John Axford hasn’t lost is his ability to miss bats. Command isn’t exactly a strength of his, but it wasn’t one of his strengths even when he was going good. I’ll cheat and tell you that Axford hasn’t lost any velocity. His stuff seems about the same, and his movement seems about the same. Look at that table and you wouldn’t conclude that a guy has completely changed, in terms of results and public opinion. But the player whose stats are shown in that table has gone from being a fan-favorite shutdown closer to a guy no one wants to see in a high-leverage situation. Or almost any situation, if we’re being honest.

That scout above said something about Axford’s release point. This isn’t perfect, but Axford’s release point data, via Brooks Baseball:

axfordrelease

That image doesn’t tell us about the specifics of Axford’s delivery, and maybe he’s lost a bit of deception. But his horizontal release point hasn’t significantly changed, and his vertical release point hasn’t significantly changed. Those images are the picture of stability, so we’d have to be dealing with something subtle. And I’ll remind you that Axford hasn’t lost his ability to miss bats. If Axford were giving away what he was going to throw, why are hitters still swinging through his pitches just as often?

Baseball Heat Maps has a useful resource. Based on Gameday information, it can spit out approximate batted-ball distances. The numbers aren’t perfect, but they don’t mean nothing, so let’s take a look at what they’re telling us. Since the start of 2012, Axford has allowed 14 home runs, and his average home runs and fly balls have traveled 294 feet. Throughout 2010, Axford allowed one home run, and his average home run and fly balls traveled 289 feet. There’s a bit of an increase there, but it isn’t massive. It’s a very small leap, that doesn’t seem to explain the profound differences in success.

In searching for explanations, I’m left with one very simple one: dumb luck. Based on a lot of the numbers, it doesn’t seem like John Axford has changed very much. His results, though, have changed quite a bit, mostly because of the home runs. How do we usually handle home-run rates? We know they take a long time to stabilize, and relievers throw even fewer innings than starters do. We remember reliever dingers more, because the stakes tend to be higher, but that doesn’t mean the numbers don’t still have to be treated responsibly.

What does luck look like? Here are three fly balls Axford allowed in his final four appearances in 2010. None of them were home runs. Axford allowed one home run all season.

Axford2.gif.opt

Axford3.gif.opt

Axford1.gif.opt

It’s such an unsatisfying conclusion, luck. And I’m not saying that’s all that’s going on. There might be something up with Axford’s delivery, and there’s probably something up with Axford’s confidence, now. But Axford probably wasn’t as good as his home-run rate before, and he’s probably not as bad as his home-run rate now. Just last year, in the first half, J.A. Happ faced 433 batters and allowed 17 home runs. In the second half, Happ faced 194 batters and allowed two home runs. In the first half, Mat Latos faced 435 batters and allowed 17 home runs. In the second half, Latos faced 423 batters and allowed eight home runs. Axford has faced 326 batters since the start of last season. He’s barely thrown 70 innings. Just because it’s more than one year doesn’t mean it’s not a small sample, and we all know how to feel about small statistical samples.

I’m not a John Axford expert, so don’t take this as gospel. Maybe Axford is really completely screwed up. But I personally don’t think he’s that different from how he used to be, and this is reflected in the hitters’ collective rate of swings and misses. If Axford were way too hittable, some of his numbers wouldn’t make him look so unhittable. It’s difficult to reconcile an excellent contact rate with the idea that hitters know what’s coming.

Remember, if John Axford is unlucky now, he was lucky before. He’s probably not as good as his career peak. But can John Axford be an effective high-leverage reliever? You can get pretty far by striking out three of every ten batters you face. I guess you can call me something of a John Axford believer. There’s gotta be at least one of us.



Print This Post



Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.


Sort by:   newest | oldest | most voted
FartBrains
Member
FartBrains
3 years 3 months ago

I think his LOB% for the last few seasons points towards luck also.

Rob
Guest
Rob
3 years 3 months ago

“But his horizontal release point hasn’t significantly changed, and his vertical release point hasn’t significantly changed.”

What would be the definition of “significantly changed?” My feeling is that even a slight change on release point would be significant. I really don’t know how to read the release point images but the horizontal one appears to be consistently changing.

I watch every game and the “scout” in me sees that the last year + he cannot put his fastball down in the zone consistently for a strike. Would that have any relationship with the seemingly shifting horizontal release point?

Nathaniel Dawson
Guest
Nathaniel Dawson
3 years 3 months ago

Any one hitter sees a certain closer very rarely, and quite often separated by a few weeks or months. It would be next to nigh impossible for a hitter to pick up a subtle shift in release point. Which doesn’t mean he isn’t tipping his pitches. It’s not always about release point, it could be he’s holding his glove up higher when he throws a curveball, or he opens his body a little more when he’s throwing a slider, or whatever. Tells.

If he’s somehow throwing his pitches differently, or tipping them in some way, how come that doesn’t show up in the stats that Jeff looked at? You’d think that something that could cause that much of a drastic difference in results should show up in most of his basic performance stats.

bgburek
Member
bgburek
3 years 3 months ago

Axford’s Batted Ball Stats
HR/FB rate FB rate
2010 2.4% 32.6%
2011 6% 35.1%
2012 19.2% 29.7%
2013 50% 66.7%

Cidron
Member
Cidron
3 years 3 months ago

not sure we can look at numbers for 2013 yet, given the far less than small sample size. but, looking at the last three, FB seems similar’ish, but, those that get into the air, are getting far more distance. One in fifty left the yard in 2010, compared to one in FIVE last year.. Thats dramatic

Nathaniel Dawson
Guest
Nathaniel Dawson
3 years 3 months ago

Not far more distance. Look above in Jeff’s article, he covers that.

It might be something weird like more distance on the balls that are hit well, and less distance on the ones that aren’t, causing more homers but not a significant change in average distance.

STEALTH
Guest
STEALTH
3 years 3 months ago

His velocity may not have been down last year, but it has been down so far this year. Nobody seems to know why that is. He says he’s not injured, and I actually believe he’s giving us the truth.

I do tend to believe he’ll be back in form at some point in or before June. Will he get his job back? That’s probably up to Jim Henderson, I guess…

Matthew Murphy
Member
3 years 3 months ago

Tremendous work, Jeff.

The_Visionary1
Member
The_Visionary1
3 years 3 months ago

Somewhat unlucky this year? yes. Lucky last year? sure. But 6 runs and 4 homers in 2.2 innings? Just because we can’t find it in the stats (yet) does not make it luck.

schuch10
Member
schuch10
3 years 3 months ago

Maybe a WHIP difference of 1.44 compared to a WHIP under 1.20 has something to do with it. Jose Valverde told me only he can play with that kind of fire. And then you throw in a bit of “bad-luck” and yeah, it is a disaster. From the games I saw, he couldn’t locate his curve in the strike zone, which made him a one pitch pony vs good teams.

Craig
Guest
Craig
3 years 3 months ago

He seems to be tipping his pitches and lost any command in his curveball so that hitters have just been sitting on a straight fastball. A couple of years ago, he’d throw his curveball for strikes in unpredictable counts and then blow his fastball by hitters who didn’t know what to expect. Also, he used to throw his slider which comes in harder than his curveball and is easier to command but obviously didn’t break as much. However, he seems to have abandoned his slider now. Without a 3rd pitch or a consistent 2nd pitch, batters just wait for the fastball which always was pretty straight. Even at 96, a straight fastball is a straight fastball, though it will still get a fair amount of whiffs due to speed alone.

I believe his biggest issue is confidence which may have lead to some slight mechanical adjustment and caused a tipping of pitches of some sort.

geefee
Guest
geefee
3 years 3 months ago

It’s not necessarily all luck (it’s actually a pet peeve lf mine when people assume that anything that happens in a small sample is without probative value) but that’s infinitely more likely than assuming that half of the fly balls he allows will being going out of the park from now on, and half of balls in play will be hits. There’s just absolutely no chance of that being the case. Hell, there’s no chance his HR/FB rate from last year will persist long term. Is John Axford suddenly a unique player in baseball history in his tendency to allow homers, while showing no other particularly negative indicators, or did he happen to allow a bunch of homers in a brief period? I know where my money is.

Here’s how you know if a move is overly reactionary: if he pitched 2.2 such innings in the middle of an otherwise normal season, almost certainly not. This obviously fluky trio of appearances happened to have come at the beginning of the season, so they’re given extra weight. And as the article alluded to, nobody would think anything of it if a starter had 2.2 such innings at any point.

I know scouting is cool again, but that pitch tipping sounds like a load. Is he just tipping the ones that go out of the ball park, or the ones that the hitters miss too? Cause they’re doing more of that than ever. Sounds like scouts bending over backward to find a satisfying explanation for what seems like a classic case of Shit Happens.

Ruki Motomiya
Member
Ruki Motomiya
3 years 3 months ago

If he is only tipping a specific type of pitch, hitters may be whiffing on the pitches he is not tipping. In addition, just because he is tipping the pitch does not necessarily mean he is tipping off location: they might have expected, say, a tipped fastball outside due to count and it comes inside, so they miss the adjustment, or the fastball is thrown further out than expected and they miss…or maybe they see a curveball coming and expect it to be out because it’s 0-2 and they figure he wants him to chase, but it hits the zone. Just because someone is tipping off doesn’t mean the hitter necessarily will always hit the tipped pitch or that every pitch is being tipped.

Flharfh
Guest
Flharfh
3 years 3 months ago

In his last game he pitched a clean inning that consisted of two warning track fly balls and a strikeout, then he came back out and gave up a monster home run. The luck theory doesn’t pass my eye test, he is straight up bad. Seems to me he can’t locate his off speed pitches and hitters are waiting for his fastball.

LionoftheSenate
Guest
LionoftheSenate
3 years 3 months ago

Why do I never see ball/strike situations much, or 1st strike % much? Esp in correlation with HR%?? Axford is erratic, from batter to batter, he is in fact consistently erratic. I think when he is getting behind now (more often than in his peak) he is basically serving up BP fastballs with little confidence….even if they are still 95MPH BP fast balls. He is lacking confidence in those key pitches where he must come in the zone.

As for tipping pitches. I’ve played with enough baseball people, and known a few scouts, to know a few baseball people are absolutely brilliant at spotting a pitcher that is tipping a pitch. Then you have the kind of tipping that goes on in the batter’s box…where a slight tweak in delivery results in easily being able to see what’s coming or at least track the ball very early. It is not every pitch, but it is often enough to get crushed every 5th guy.

Luck is weak. We know data is limited in short samples. That’s why you need to open your eyes and actually watch a game now and then.

geefee
Guest
geefee
3 years 3 months ago

Is luck “weak,” or does it not exist? I’m not sure which of those is more wrong.

Luck is boring/for pussies, so obviously it’s not luck -> something must be awry, and that something is ascertainable just by watching. I don’t know about that reasoning. We all know any observations won’t be objective, if you’re looking for an explanation, you’re going to find it (prime example: lack of confidence). Sounds like a recipe for misdiagnosis.

LionoftheSenate
Guest
LionoftheSenate
3 years 3 months ago

Oh and Axford’s velocity is down 2 MPH, at least. How was this missed? Also, I think it would be better to break out the release point charts by pitch type. Don’t lump all pitches on same chart, really not easy to read or analyze. Try FB and CB. That’s most of what he throws.

gryfyn1
Member
gryfyn1
3 years 3 months ago

maybe Axeford just isn’t really any good; maybe 2010-11 were the outlier just the equivalent of a hot 1st half of the year for a starter. And prehaps the conversation should be how he managed to be good in those years.

Bob
Guest
Bob
3 years 3 months ago

Go to Brooks Baseball, and look at John Axford’s career trends with his pitches. He started throwing his curve ball with more movement in 2012, as compared to earlier in his career. That corresponds with an 8% increase (from 46% to 54%) in the frequency of curves thrown for balls.

Axford throws 3 pitches – fastball, curve, and slider. Since being promoted to closer, he has increased his use of the fastball (65% -> 71%), decreased his use of the slider (18% -> 9%), and kept the curve relatively constant (17% -> 18%).

Now we see three trends. The first is that he is throwing more fastballs. The second is that he is throwing fewer sliders. The third is that he is throwing fewer curves for strikes. It makes sense to me that an MLB caliber hitter would be able to capitalize on these trends.

Bob
Guest
Bob
3 years 3 months ago

Also look at his batted ball profiles, in addition to contact%.

2010-2011: 49% GB; 34% FB; 17% LD
2012-2013: 43% GB; 32% FB; 25% LD

He is allowing fewer GB/FB, and more LD. LD result in hits more frequently than GB or FB. Which also seems to correspond with hitters being able to capitalize on his trends.

vivaelpujols
Guest
vivaelpujols
3 years 3 months ago

In other words, scouts are still full of shit half the time. The “tipping pitches” line has got to have been used 100 times by now.

Bob
Guest
Bob
3 years 3 months ago

I don’t think it is “tipping pitches” as much as being more predictable.

ralf
Guest
ralf
3 years 3 months ago

I’d like to see someone dig through pitchFX and look for changes in location. Seems like he’s been missing in the zone more often. He still gets swings and misses on the high FB and low curve, but just watching the games it looks like he’s painting the corners less often and putting more pitches in hittable spots.

Related: has anyone done a study on the many relievers who have had short, high peaks and steep, fast declines without getting seriously injured? I’d like to see what, if anything, they have in common with each other. Even when Axford was doing well, a lot of Brewers fans had a feeling he’d be the next Turnbow. And if Henderson gets the closer’s job I can see him going down the same road.

Jack Weiland
Guest
Jack Weiland
3 years 3 months ago

Warning: I’ve done no research to back up what I’m about to say.

I’m struck by how many areas Axford has trended in the wrong direction, in a small way, and they all seem to be passed off as “Oh it’s not a huge decrease.” I’ll agree that luck is likely at play here in some fashion, be it small or large. But it’s not a requirement to have huge drops in key areas for performance to tank, is it? Isn’t it possible that those small differences, especially the cumulative effect of many of them, are partially responsible for the change in results?

vivalajeter
Guest
vivalajeter
3 years 3 months ago

On a scale of One to Halladay? I’d give him a six.

phoenix2042
Guest
phoenix2042
3 years 3 months ago

Can we see his edge percent (you know how that stat got introduced in the lincecum article a while back?) and see if maybe he’s missing over the plate more often?

Thomas Vockrodt
Guest
Thomas Vockrodt
3 years 3 months ago

Tipping pitches is no big deal. Mariano Rivera has one pitch. If the quality of the pitch is good, it is more deflating knowing what is coming and not being able to hit it.

Axford has terrible stuff and poor body language which tips off the quality of his pitches.

Frank Q
Guest
Frank Q
3 years 3 months ago

“Axford has terrible stuff… which tips off the quality of his pitches.”

Are you blowing off steam or else what are you saying? Because 1 pitcher can throw basically one pitch (although he always used a four-seamer and now incorporates a two-seamer) that means that tipping off pitches isn’t significant for all pitchers? And Axford has terrible enough stuff to not have a career K rate of over 11K / 9 IP. And his “terrible stuff” plays out worse which is why he can continue to be post one of the better K rates in the NL?

valley_man0505
Guest
valley_man0505
3 years 3 months ago

I’m assuming the flyball GIFs from 2010 are to show that even when he was good, he still gave up long fly balls. You also say that the change in average batted ball distance from 289 ft throughout 2010 to 294 ft since 2012 is a “very small leap”. It may be a “very small leap” for someone going from 260 to 265, but when you are always living at the warning track, 5 feet is a lot. Add 5 feet of distance to each of the fly balls in the GIFs and tell me that 5 feet is a “very small leap”.

Also, I agree with what Jack says above, there may not be one thing that is dramatically different, but there does seem to be a lot of small things going in the wrong direction. 1 x 100 = 100, but so does 10 x 10.

papasmurf
Guest
papasmurf
3 years 3 months ago

Are there any data on pitch location or pitch movement? I think the posters’ comments about his loss of curveball command is a far better explanation than “luck.”

All strikes aren’t the same. A quality strike on/near the corners of the strike zone is more likely to be hit for an out, while those nearer to the middle tend to get hit harder and more likely for hits. A fastball that was half an inch off the sweet spot of the bat may have made it to the warning track. But you move that half an inch toward the heart of the plate and maybe you get 20 extra feet and a homer.

wpDiscuz